Urosepsis after transrectal ultrasonography-guided prostate biopsy: reaudit following a shortened antibiotic prophylaxis regimen

Abstract

Sir, We read with interest this journal’s recent Leading article concerning the national burden of infections after transrectal ultrasonographyguided prostate biopsy (TRUSgpb) and the need for urgent action to reduce infectious complications. The authors discuss a number of possible strategies, including alterations in antimicrobial prophylaxis. We recently performed an audit cycle to specifically assess one such strategy. In 2009, we published our Trust’s post-TRUSgpb infection data for theyear October 2007–September 2008. Although the incidence of urosepsis [7/256 (2.7%) biopsies, including 4 (1.6%) bacteraemias] was not out of keeping with other reports, our prime concern was the multiresistant nature of the pathogens, with 100% ciprofloxacin resistance and 57% extended-spectrum b-lactamase (ESBL) production. During this time, the antibiotic prophylaxis regimen comprised ciprofloxacin and metronidazole for 1 day prior to biopsy and 2 days post-biopsy. To lessen effects on the bowel flora before the procedure, and to align with protocols used in other centres, prophylaxis was altered to begin just 2 h before biopsy (a reduction of two doses of ciprofloxacin and three of metronidazole). Our infection rate was then reaudited over 2 years. Some 547 TRUSgpbs were performed between 1 July 2009 and 30 June 2011 at Wycombe Hospital. All patients were prescribed 500 mg of ciprofloxacin every 12 h plus 400 mg of metronidazole every 8 h, starting 2 h prior to the procedure and continuing until 48 h post-procedure. Subsequent cases of urosepsis (bacteraemia plus those with positive urine cultures requiring readmission) were identified and their notes reviewed. Table 1 combines the results from both audits. In the second study period, urosepsis was diagnosed following 4 of 547 procedures (0.73%), with three confirmed bacteraemic episodes. In comparison with the initial audit, the bacteraemia rate was not significantly reduced (3/547 versus 4/256, P1⁄40.218, Fisher’s exact test), but the decrease in the overall urosepsis rate reached statistical significance (4/547 versus 7/256, P1⁄40.044). In the 2009–11 period, all four cases were caused by Escherichia coli, with all strains being ciprofloxacin resistant. Only one was an ESBL producer, but two were also gentamicin resistant. As described by Batura and Gopal Rao, the increased incidence of urosepsis post-TRUSgpb in recent years has fuelled interest in how to improve antibiotic prophylaxis, amongst other strategies, to reduce infectious complications. The Cochrane Collaboration systematic review of 2011 concluded that antibacterial prophylaxis significantly reduced the incidence of all infectious outcomes studied (fever, bacteraemia, bacteriuria, urinary tract infection and hospitalization). Most available data supported the use of fluoroquinolones as the most suitable prophylactic agent. Interestingly, when the duration of treatment was analysed, no significant benefit or detriment was attributed to long-course (3 day) over short (1 day/one dose) regimens. Furthermore, the equivalence of single-dose ciprofloxacin to that of longer regimens has more recently been reported in a large US study. The data presented here should be viewed with this in mind. It is possible that the change in the prophylaxis start time from 24 to 2 h pre-biopsy reduced the antimicrobial effect on the bowel flora, with less time for ciprofloxacin-resistant organisms to become predominant and, hence, fewer men developed urosepsis secondary to such pathogens. It may also be that by timing a dose of ciprofloxacin 2 h prior to the procedure, higher and more consistent antibiotic levels in serum and prostatic tissue resulted than with the former dosingschedule.However,prophylactic regimensof≥48 h,as administered here, may be considered unnecessarily long. Indeed, a single dose of a fluoroquinolone is currently advocated as suitable prophylaxis by the European Association of Urology (EAU). Ciprofloxacin-resistant E. coli is by far the predominant pathogen in post-TRUSgpb sepsis. This is not surprising, given that ciprofloxacin is the main prophylactic agent used and a significant proportion of patients have received ciprofloxacin to treat urinary infections associated with their prostatic problems. Recent studies have shown that 10%–20% of men undergoing TRUSgpb harbour quinolone-resistant organisms in their bowel prior to the administration of prophylactic antibiotics and a link between such colonization and post-TRUSgpb infection has been demonstrated. The significant finding of coincident ESBL production and gentamicin resistance in up to 75% and 33% of causative organisms isolated, respectively, has also been reported. Total numbers in our repeated audit are still too small to usefully compare the occurrence of these two latter characteristics. However, analysis of plasmids has shown the increasing presence of multiple genes encoding ESBL enzymes and other mechanisms of resistance to important antibiotics, including gentamicin. In 2009, we hypothesized that rectal screening for fluoroquinolone-resistant and ESBL-producing bacteria shortly before TRUSgpb, with antibiotic prophylaxis tailored to the results, may be more effective at reducing post-biopsy urosepsis than blanket regimen changes. Although financial constraints have so far prevented such screening procedures being introduced in our Research letters